Artificial Intelligence & Machine Learning


Sean Davis, MD, PhD

Department of Biomedical Informatics
University of Colorado | Anschutz Medical Campus

Tuesday, July 11, 2023

Overview

  • Brief history and background
  • What are these things?
  • Machine learning
  • Biases and ethics

Brief History

The hype cycle

The Gartner Hype Cycle.

The hype cycle (circa 2016)

The Gartner Hype Cycle.

What is Artificial Intelligence?

The theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.

What is Machine Learning?

The study of computer algorithms that improve automatically through experience. It is seen as a subset of artificial intelligence. Machine learning systems give the computer the ability to learn without being explicitly programmed rules.

What is Deep Learning?

Machine learning algorithms that are inspired by the structure and function of the brain. Deep learning is a subset of machine learning in artificial intelligence that has networks capable of learning unsupervised from data that is often unstructured (i.e., text or images).

AI vs. ML vs. Deep Learning

Artificial
Intelligence

Machine
Learning

Deep
Learning

AI vs. ML vs. Deep Learning

Artificial
Intelligence

Machine
Learning

Deep
Learning

Applications in healthcare

The data types considered in the artificial intelligence artificial (AI) literature. The comparison is obtained through searching the diagnosis techniques in the AI literature on the PubMed database(Jiang et al. 2017, fig. 1).

Applications in healthcare

The leading 10 disease types considered in the artificial intelligence (AI) literature. The first vocabularies in the disease names are displayed. The comparison is obtained through searching the disease types in the AI literature on PubMed (Jiang et al. 2017, fig. 3).

Applications in healthcare

from (Challen et al. 2019)

Machine Learning

Classes of Machine Learning

Broad classes of machine learning.

A map of machine learning approaches

Supervised Learning

Supervised Learning

Supervised learning.

Supervised Learning

Supervised learning.

Supervised Learning

Supervised learning.

Classification

Regression

Application of regression to predict eutrophil count using the signatures learned from public databases(Oh et al. 2022, fig. 4)

Developing a model

  • Develop question!!!

  • Collect data

  • Split data into training and test

  • Train model

    • estimating model parameters (i.e. training models)
    • determining the values of tuning parameters that cannot be directly calculated from the data
  • Test model

    • calculating the performance of the final model that will generalize to new data
  • Possibly validate model

  • Deploy

Algorithms for Supervised Learning

Linear regression.

Algorithms for Supervised Learning

Anscombe’s Quartet

Algorithms for Supervised Learning

Classification and Regression Trees (CART).

Algorithms for Supervised Learning

K-nearest neighbor (kNN) algorithm.

Algorithms for Supervised Learning

Random Forests.

Algorithms for Supervised Learning

Deep learning.

Applying Supervised Learning Algorithms

The mlr3 ecosystem in R.

Unsupervised Learning

Unsupervised learning.

Clustering

Gene expression measurements.

Classes of unsupervised learning algorithms

Clustering.

Dimensionality reduction.

Clustering

Thinking about similarities and differences

Clustering

Algorithm for hierarchical clustering

Clustering

Algorithm for hierarchical clustering

Clustering

Algorithm for hierarchical clustering

Dimensionality Reduction

Schematic PCA.

Dimensionality Reduction

Using dimensionality reduction to explore 22,000 dimensions of gene expression data on 280 samples.

Machine Learning Review

  • Supervised Learning
    • Classification
    • Regression
  • Unsupervised Learning
    • Clustering
    • Dimensionality Reduction

Resources

Ethics in AI & ML

Ethical challenges in AI & ML

How do we address ethical and moral decisionmaking for individuals, groups, and society?

Biases in AI/ML

Types and sources of bias in big data and AI/ML applications in healthcare (Norori et al. 2021). Bias in the medical field can be dissected along three directions: data-driven, algorithmic, and human.

Addressing Bias

Open science practices can assist in moving toward fairness in AI for health care (Norori et al. 2021). These include:

  1. participant-centered development of AI algorithms and participatory science;
  2. responsible data sharing and inclusive data standards to support interoperability;
  3. code sharing, including sharing of AI algorithms that can synthesize underrepresented data to address bias and improve health outcomes.

Each of these is harder and more expensive than it sounds.

Questions and Discussion

References

Challen, Robert, Joshua Denny, Martin Pitt, Luke Gompels, Tom Edwards, and Krasimira Tsaneva-Atanasova. 2019. “Artificial Intelligence, Bias and Clinical Safety.” BMJ Quality & Safety 28 (3): 231–37.
Jiang, Fei, Yong Jiang, Hui Zhi, Yi Dong, Hao Li, Sufeng Ma, Yilong Wang, Qiang Dong, Haipeng Shen, and Yongjun Wang. 2017. “Artificial Intelligence in Healthcare: Past, Present and Future.” Stroke and Vascular Neurology 2 (4): 230–43. https://doi.org/10.1136/svn-2017-000101.
Norori, Natalia, Qiyang Hu, Florence Marcelle Aellen, Francesca Dalia Faraci, and Athina Tzovara. 2021. “Addressing Bias in Big Data and AI for Health Care: A Call for Open Science.” Patterns (New York, N.Y.) 2 (10): 100347. https://doi.org/10.1016/j.patter.2021.100347.
Oh, Sehyun, Ludwig Geistlinger, Marcel Ramos, Daniel Blankenberg, Marius van den Beek, Jaclyn N Taroni, Vincent J Carey, Casey S Greene, Levi Waldron, and Sean Davis. 2022. GenomicSuperSignature Facilitates Interpretation of RNA-seq Experiments Through Robust, Efficient Comparison to Public Databases.” Nature Communications 13 (1): 3695. https://doi.org/10.1038/s41467-022-31411-3.